Introduction

Kawasaki syndrome (KS) is an acute self-limited medium size vasculitis, principally affecting coronary arteries (CA) and children less than 5 years of age [1]. KS is one of the most common cause of acquired heart disease in children, and it is most frequent in Asian countries, particularly in Japan, where annual incidence of 308 cases per 100,000 children < 5 years of age has been detected [2, 3]. Conversely, in Europe the incidence ranged between 4.9 and 15.2 per 100,000 children <5 years of age [4]. It is characterized by persistent fever, exanthema, conjunctivitis, changes in the extremities, erythema of oral mucosa and lips and cervical lymphadenopathy along with coronary abnormalities [1]. Rarely, macrophage activation syndrome may complicate the clinical course of KS [5]. The aetiology is still unknown, even though many studies are still investigating the possible triggers of this vasculitis. Identification of risk factors for coronary artery lesions (CAL) in KS represents the current purpose of medical community to early detect patients needing a tailored therapy [1].

Objective

To describe epidemiological, clinical e laboratory characteristic of a mono-centric KS cohort, followed at Rheumatology and Cardiology Units between February 1982 and August 2018.

Materials and methods

This is a monocentric, retrospective, observational study involving patients with a diagnosis of KS between February 1982 and August 2018 at Rheumatology and Cardiology Unit of Anna Meyer Children’s University Hospital. Complete and incomplete KD were defined according to the American Heart Association definition (1). In a customized database, clinical, laboratory and echocardiographic data have been extracted from medical records, stratifying subjects with no CAL, coronary ectasia (CE) and presence of coronary aneurisms (CAn) according to American Heart Association recommendation 2017 [1]. All patients included underwent to echocardiographic evaluation. Resistance to treatment was defined as persistence of fever 48 h by the end of IVIG infusion. Regardless CAL, patients underwent scheduled cardiologic follow-up until 6–8 weeks from the onset, thereafter at 6 months, 1 year and at 8 years, including ergometric test with tapis roulant (treadmill) according to Bruce protocol modified [6]. CAL patients received additional cardiac evaluation on clinical need. Patients with < 1-year follow-up were excluded from analysis regarding data of follow-up. Statistical analysis was performed using SPSS statistics version 25.0. Data are expressed as frequencies and total records, median, range or mean and standard deviation (SD). Data have been assessed by analysis of variance with Tukey, LSD and Bonferroni corrections, and r-Spearman and χ-square tests.

As per Anna Meyer Children’s Hospital Ethical Committee, ethics board approval is not required for retrospective observational study.

Results

Between February 1982 and August 2018, 361 children have been diagnosed with KS, 47 patients between the 1982 and 1991 (follow-up range 1–36 years), 88 between the 1992 and 2001 (follow-up range 2–26 years), 141 between the 2002 and 2011 (follow-up range 1–16 years) and 85 between the 2012 and August 2018 (follow-up range 1 month −6 years). Among these, 219 patients were male (60.6%), with a male/female ratio of 1.5:1. They aged 1 month to 24.4 years, with a median age at disease onset of 2.1 years. Three hundred and twenty-five out of 361(90%) children developed the disease before 5 years, 94/361 (26%); before 1 year, 32/361 (8.9%); and before 6 months. All patients presented persistent fever (100%), 244 rash (91.4%), 229 conjunctivitis (85.8%), 224 mucositis (83.9%), 196 extremities abnormalities (73.4%) and 165 lymphadenopathies (61.8%). Patients younger than 6 months of age in the 68.75% (22/32) of case had an incomplete form of KS, versus the 46.64% (132/283) of patients aged 6 months-5 years, versus the 33.33% (15/45) of patients older than 5 years (χ2 9.46, p value 0.008). The median fever duration was 7 days (range 1–25 days). CAL was detected in 73/361 patients (20.2%): 58/361 had CE (16%), and 15/361 had CAn (4.15%).

In the different subgroups, we observed a median age at disease onset of 2 years and 1 month in No-CAL, 2 years and 5 months in CE, 5 months in CAn, a median duration of fever of 7 days in No-CAL, 8 days in CE, 11 in CAn, an average value of CRP of 6.81 mg/dl in No-CAL, 8.38 mg/dl mg/dl in CE, 14.15 mg/dl in CAn, a day of first dose of IVIG of 8th day in no-CAL, 7th day in CE and 11th in Can (see Table 1).

Table 1 Clinical and laboratory characteristic in the 3 different groups of 361 Kawasaki Syndrome children
Full size table

Among 169 patients with incomplete KS, children younger than 6 months developed CAL more frequently than children aged between 6 months and 5 years and children aged more than 5 years (7/22 vs 11/132 vs 3/15 (χ2 10.42, p value 0.005).

The different groups showed significant differences regarding age at disease onset (F = 2.77, p = 0.025), fever duration (F = 16.32, p < 0.0001), CRP (F = 6.94, p = 0.001) and day of first IVIG administration (F = 7.963 p < 0.0001) (Fig. 1).

Fig. 1
figure 1

Figure reports significant differences regarding a days of fever, b CRP values and c age at disease onset in Kawasaki syndrome children without coronary artery lesions, coronary ectasia and coronary aneurisms. In detail, section a highlights longer duration of fever in patients with coronary aneurism (p < 0.0001), section b highlights lower values of CRP in patients without coronary artery lesion (p 0.001), section c highlights younger age in the group of coronary aneurism (p 0.025). no-CAL, non-Coronary artery lesion; CE, coronary ectasia and CAn, presence of coronary aneurism

Full size image

A significant correlation between CAL and disease onset < 6 months (ρ = 0.137, p = 0.009), the need to administer second IVIG dose (ρ = 0.305, p < 0.001) and male gender (ρ = − 0.109, p = 0.038) has been detected. No significative associations were evaluated among the presence/absence of CAL and conjunctivitis, cheilitis, rash, lymphadenopathies, hyponatremia and white blood count > 12.000 cell/mm3.

Until the 1986 (8 children), corticosteroids associated with high dose of aspirin was the main therapy. From the 1986 to 2005 (205 children), IVIGs, one or two doses, along with aspirin have been used, an additional third course of IVIG has been provided in 35 children, before receiving corticosteroids in case of persistent disease (10). From 2005 (148 children), after the failure of a second dose of IVIG, 36 patients received 3 boluses of corticosteroid. After 2012, Anakinra (2) and Infliximab (1) have been administered in severe/refractory forms not responding to steroid treatment.

Data of patients with at least one-year follow-up were available for 327 children. Median long-term follow-up was 10.2 years (range 1–36 years). At the last available follow-up, 261/327 no CAL patients and 53/327 CE patients had normal ECG, echocardiography and ergometric test (performed in 177/261 patients with no CAL and in 39/53 with CE). Conversely, among 13/327 CAn patients, 8/13 persistent CAn at echocardiography, one displayed ECG repolarization alterations. Nine out of 13 children underwent an ergometric test: one child showed ST depression at the peak of effort.

During this long-term follow-up, none of our patients developed additional systemic signs or features of others autoinflammatory diseases.

Discussion and conclusion

In this report, we have described the features of a cohort of KS in a tertiary Italian centre. Compared with literature data [7], we observed a lower age at disease onset: 90% patients developed KS before 5 years. However, the percentage of CAL (20.2%) is consistent with European data [4]. Furthermore, the subgroup with CAL showed a lower age at disease onset probably due to unusual/incomplete presentation and paucity of clinical signs, as previously reported. It is unclear if the increase risk of CAL is the consequence of the delayed diagnosis or to the age of the children [8].

Our long-term follow-up in a large, monocentric cohort reports possible CAL risk factors such as age at disease onset, fever duration, CRP level and day of first IVIG administration according to current data [9]. These parameters are part of different scores developed to predict CAL and IVIG resistance, but unfortunately there is no agreement about their applicability across populations with different genetic background [10]. Our study confirms, over 36 years of observation in real life, the benign course of KS without CAL after 6–8 weeks from the disease onset [1, 11, 12]. Additionally, it reports, in long-term follow-up, the favourable outcome of children with CAn: 40% no longer presented CAL and, significantly, only 7.69% showed abnormal function tests.

Nonetheless the potential interpretation bias due to be a monocentric, retrospective chart review, these data outlines the benign course of KS over time, in no CAL subjects as well as in children with CAn.

According to recent guidelines, stopping cardiologic assessment in no risk patients might result timesaving and able to reduce emotional discomfort in children and their families as evaluated by de Ferranti and colleagues in their recent study (12). Our long-term data confirms the benign course of KS and should prompts physicians in tailoring diagnostic and therapeutic efforts in selected KS children.